I recently asked a question that was - long story short - answered by:
There is an
app'package for that!
This package indeed fully addresses my needs, but I was not aware of it (even after web search). In my question, I am thus showing my effort to develop this functionality (since I didn't succeed in finding the package by myself, so assuming it didn't exist). I explore various ways to solve my issue: some are successful but partial, some are unwieldy, and some are plain cul-de-sac.
So I think this question (and my attempts/partial solutions) still has some value for future user, who want - as I was - to use environments developed in the tufte
class without using this class, and without knowing the existence of the sidenote
package. However, expressed as it is now, this trials are "noise" between a simple and clear question, and its simple and clear answer.
General question: Once one got an answer to his/her question, what should be trimmed in the original post in order to make it as valuable as possible for future users?
(That means in my particular case: what should I trim in my original post -- that is, I admit, a bit dense -- in order to make it valuable?)
E.g., Should I remove all the work I did (trying to replicate tufte
functionality) showing my efforts in my MWE, but that is now useless? It would make the question cleaner, but will lose content and potential "entry points" for people having same issue I had.
Or should I post it as an answer (so that this work is not lost) but with a disclaimer saying that it's for archive only (and that DG' gave the satisfactory answer)?
former code
can be seen in the edit history.